Move your team from Slack AI to Dust.

Both platforms let AI agents call into third-party systems as part of a workflow. In Dust, agents are given a curated set of tools (web search, Notion/Slack/GitHub/Salesforce actions, MCP servers) that the agent selects and invokes autonomously during a conversation, guided by its instructions. In Slack AI, third-party AI agents (Salesforce Agentforce, Asana, Writer, Cohere, Workday, Box, Adobe Express) are installed as Slack apps and invoked by @-mentioning them in a channel or DM. To migrate from Dust to Slack AI: (1) list every third-party tool your Dust agents invoke (Salesforce, GitHub, Zendesk, etc.); (2) check whether an equivalent Slack AI agent app exists in the Slack App Directory for each tool; (3) install those Slack agent apps and grant them channel access; (4) retrain users to @-mention the relevant agent instead of prompting the Dust agent. To migrate from Slack AI to Dust: (1) note which @-agent integrations your team relies on; (2) connect the same underlying systems (Salesforce, GitHub, etc.) as Dust data sources or action tools; (3) write a Dust agent whose instructions replicate the task the Slack agent handled.

Both products support automating repetitive AI-assisted tasks, but through different mechanisms. Dust exposes a REST API (Bearer token auth) so that scripts, Zapier/Make/n8n automations, and Slack-based triggers can call any Dust agent programmatically; every such call is metered against workspace credits. Slack AI's Workflow Builder lets users describe a desired automation in plain language and have Slack AI generate the workflow steps, which then run inside Slack and connect to external apps via Slack's built-in integrations. To move automated workflows from Dust to Slack AI: (1) list each programmatic Dust agent invocation (Zapier zap, n8n flow, Slack shortcut) and the action it performs; (2) open Workflow Builder in Slack and use AI workflow generation to recreate the trigger-step logic in Slack-native terms; (3) connect external apps (Google Drive, Salesforce) via Slack integrations to replace Dust data-source calls. To move from Slack AI to Dust: (1) export or document your Slack Workflow Builder flows; (2) create matching Dust agents with equivalent instructions and data connections; (3) call those agents via the Dust API from the same trigger events (Slack message posted, form submitted, etc.) using Zapier, n8n, or direct API.

Both products offer a conversational AI agent that team members interact with via chat to draft content, answer questions, and assist with recurring tasks. On Dust, custom agents are built in the no-code agent builder: write natural-language instructions, attach knowledge sources (Notion, Google Drive, etc.), and choose a model. The resulting agent is shared across the workspace. On Slack AI, Slackbot serves as the personal assistant accessed via DM — it draws on the user's own Slack history, files, and projects and does not require any build step. To move from Dust to Slack AI: (1) identify the agent instructions and knowledge sources used in Dust; (2) ensure the underlying data (docs, briefs, specs) is searchable inside Slack or in a connected app (Google Drive, Salesforce); (3) direct users to Slackbot DMs for drafting and Q&A tasks. To move from Slack AI to Dust: (1) export any key context from Slack (pinned messages, canvases, file exports); (2) create a Dust agent using those exports as data-source connections; (3) write instructions that match the prompting patterns users relied on in Slackbot.

Both products let users ask natural-language questions answered by the organization's data, but they source that data differently. Dust data connections sync content from external apps (Google Drive, Notion, Confluence, GitHub, Snowflake, HubSpot, Zendesk, Gong, and others) into Dust's own index; agents search this index using RAG to produce answers. Slack AI Search queries the Slack message archive plus natively connected apps (Google Drive, Salesforce on Enterprise+). To move from Dust to Slack AI: (1) map each Dust connector to its equivalent Slack integration (Google Drive, Salesforce); (2) ensure key knowledge documents are shared in Slack channels or connected via the Slack App Directory; (3) train users to type questions into Slack Search rather than a Dust agent conversation. To move from Slack AI to Dust: (1) identify the connected apps that feed Slack's AI Search (Google Drive, Salesforce, etc.); (2) connect those same apps as Dust data sources via the Dust Connections admin panel; (3) verify permission scopes match so that the right users can access the synced data.

Both products surface AI assistance inside Slack channels, but they handle conversation catch-up differently. Dust agents are accessible natively in Slack — users @-mention a Dust agent in any channel or DM, and the agent answers questions, summarizes context, or takes actions by reading connected knowledge sources. Slack AI provides built-in 'Summarize' buttons on channels and threads that generate a recap without leaving Slack or mentioning an external agent. To move catch-up workflows from Dust to Slack AI: (1) confirm the team is on Business+ or Enterprise+; (2) enable Slack AI in workspace settings; (3) train users to click 'Summarize' in a channel header or thread instead of @-mentioning the Dust agent with a 'what did I miss?' prompt. To move from Slack AI to Dust: (1) connect the Slack integration in Dust (available on Pro and Enterprise); (2) create a Dust agent with instructions to summarize Slack channel content; (3) invite the Dust Slackbot to the relevant channels and show users how to @-mention it.